Systematic examination of T cell responses to SARS-CoV-2 versus influenza virus reveals distinct inflammatory profile

31 32 There is a pressing need for an in -depth understanding of immunity to SARS-CoV-2. Here we 33 investigated T cell recall responses to fully glycosylated Spike trimer, recombinant N protein as 34 well as to S, N, M and E peptide pools in the early convalescent phase. All subjects showed SARS-35 CoV-2-specific T cell responses to at least one antigen. SARS-CoV-2-specific CD4+ T cells were 36 primarily of the central memory phenotype and exhibited a lower IFN-g to TNF-a ratio compared 37 to influenza-specific responses of the same donors, independent of disease severity. SARS-CoV-38 2-specific T cells were less multifunctional than influenza -specific T cells, particularly in severe 39 cases, potentially suggesting exhaustion. High IL -10 production was noted in response to N 40 protein, possibly contributing to immunosuppression, with potential implications for vaccine 41 design. We observed granzyme B +/IFN-g+ CD4+ and CD8+ proliferative responses to peptide 42 pools in most individuals, with CD4+ responses predominating over CD8+ responses. Peripheral 43 T follicular helper responses to S or N strongly correlated with serum neutralization assays as well 44 as RBD-specific IgA. Overall, T cell responses to SARS-CoV-2 are robust, however, CD4+ Th1 45 responses predominate over CD8 + responses and are more inflammatory with a weaker Tfh 46 response than influenza-specific CD4+ responses, potentially contributing to COVID-19 disease. 47 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprintthis version posted September 1, 2020. ; https://doi.org/10.1101/2020.08.27.20183319doi: medRxiv preprint 3 The disease COVID-19, caused by the novel coronavirus (CoV), SARS-CoV-2, emerged in China 48 in late 2019 and is currently causing a devastating pandemic (1-3). Despite the severity of the 49 disease in some individuals, the vast majority of infected people recover, indicating that they have 50 made an effective immune response that clears the virus. Moreover, studies in rhesus macaques 51 demonstrate that SARS-CoV-2 induces protective immunity against rechallenge at least out to 35 52 days (4, 5). Adaptive immunity, mounted by T and B lymphocytes, is critical for clearance of viral 53 infections and for protec tion against reinfection. Most studies to date show that people infected 54 with SARS -CoV-2 produce Spike (S) and receptor binding domain (RBD) specific-IgG and 55 neutralizing antibodies within two to four weeks of infection (6-13). Although some studies have 56 suggested that antibody responses of people with mild or no symptoms can fall off rapidly (7, 14, 57 15), other studies suggest IgG responses are relatively stable over the first 3-4 months, with peak 58 responses followed by a gr adual decline as observed in a normal IgG response (13, 16, 17) . In 59 contrast, IgA responses to SARS -CoV-2 start early and decay rapidly (17). In the absence of 60 complete virus neutralization, T cells are critical for eliminating virus -infected cells. Moreover, 61 CD4+ T cell responses , and in particular T follicular helper (Tfh) responses , are critical for 62 generation of high affinity long -lived antibody responses (18). Follow-up studies of the SARS -63 CoV-1 outbreak in 2003 showed that antibody responses fell off substantially between 3 and 5 64 years in most individuals (19), whereas T cell responses could be detected for more than 11 years 65 (20). Moreover, nucleocapsid (N)-reactive T cells in SARS-CoV-1 recovered patients at 17 years 66 post-infection showed substantial cross-reactivity to SARS-CoV-2 N peptides (21). Thus, T cells 67 likely represent an important part of protective immunity to SARS-CoV-2. 68 69 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprintthis version posted September 1, 2020. ; https://doi.org/10.1101/2020.08.27.20183319doi: medRxiv preprint 4 Several studies have examined T cell responses to SARS -CoV-2 with most studies using 70 restimulation with overlapping peptide pools from several SARS-CoV-2 open reading frames (21-71 28). Responses to restimulation with intact N, S-RBD domain and protease proteins have also been 72 reported (11). The studies to date have used a variety of readouts to determine T cell spe cificity 73 including activation markers, intracellular cytokine production, IFN-g EliSpot or measurements of 74 cytokines in the supernatants by multiplex assays. In general, the majority of confirmed SARS-75 CoV-2 cases have shown CD4+ and CD8+ T cell responses to SARS-CoV-2 antigens in the acute 76 and early convalescent phase, dominated by a Th1 response with some studies also reporting Th2 77 or Th17 responses (reviewed in (29)). CD8+ T cell responses have also been detected in the 78 majority of, but not all, donors. There is also evidence of cross-reactive T cells in 20-50% of donors 79 who donated blood pre -pandemic. The se cross -reactive responses are to peptides conserved 80 between seasonal coronaviruses and SARS-CoV-2 (21, 22, 28, 30, 31). 81 82 Given the consistent findings of Th1 and CD8+ T cell responses in the acute and early convalescent 83 stage of SARS-CoV2, often with the strongest responses detected in the more severe cases, it is 84 not yet clear why the immune system fails to rapidly control the virus in some patients. Here we 85 undertook a systematic functional examination of T cell responses to SARS-CoV-2 in a cohort of 86 13 SARS-COV-2 recovered individuals with a range of diseas e severity who provided 87 leukapheresis samples in the early convalescent phase (4-12 weeks). Specifically, we examined T 88 cell phenotype, cytokine production, and proliferation to SARS-CoV-2 proteins and peptides and 89 compared them to seasonal influenza responses. We also identified peripheral T follicular (pTfh) 90 IL-2 producing CCR7+CXCR5+ cells in response to SARS-CoV-2 antigens in some donors and 91 found that the frequency of these cells strongly correlated with serum neutralization assays and 92 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprintthis version posted September 1, 2020. ; https://doi.org/10.1101/2020.08.27.20183319doi: medRxiv preprint 5 RBD-specific IgA but were less frequent than those observed in response to influenza . Our study 93 reveals new insights into the recall response of SARS -Cov-2 in the early convalescent phase, 94 highlighting that SARS-CoV-2-specific CD4+ T cell responses are more inflammatory and show 95 a weaker pTfh response than influenza A-specific CD4+ recall responses within the same donors. 96 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprintthis version posted September 1, 2020. ; https://doi.org/10.1101/2020.08.27.20183319doi: medRxiv preprint 6

97 Patient characteristics 98 Thirteen COVID-19 convalescent donors who had recently tested positive for SARS -CoV-2 by 99 PCR, and a single SARS -CoV-1 patient from 2003 were consented for leukapheresis to obtain 100 plasma and PBMC (Table 1). Samples for SARS-CoV-2 convalescent individuals were collected 101 from 27 days to 90 days post onset of sym ptoms. Disease severity ranged from asymptomatic, 102 mild (non-hospitalized), moderate (hospitalized not ICU) to severe (ICU). The average age was 103 53 (range 31-72), and 8 out of 13 were male (Table I). 104 105 Ex vivo i ntracellular cytokine responses to Spike and N proteins in convalescent COVID-19 106 patients 107 To date, most studies have used overlapping peptide pools to assess antigen -specific T cell 108 responses to SARS-CoV-2. Here we used intact glycosylated S from SARS-CoV-2 and seasonal 109 human CoV-OC43 (OC43), as well as recombinant E. coli expressed SARS-CoV-2 N to determine 110 how T cells respond functionally to SARS-CoV-2 under conditions where antigen processing is 111 required. Intracellular cytokine staining (ICC) of ex vivo PBMC was conducted to determine the 112 frequency of IFN-g, TNF-α and IL-2-producing cells, with the gating strategy shown in Figure 113 S1A. Following 18hrs of stimulation, with Golgi Stop and GolgiPlug added for the last 6 hrs, 114 SARS-CoV-2 S-specific CD4+ T cells were detected in 54% of donors based on IFN-g production, 115 75% of donors based on TNF-α production and 85% of donors based on IL -2 production. N -116 specific CD4+ T cells were detected in 38% of donors based on IFN-g, 58% based on TNF-α, and 117 54% of donors based on IL -2 production (Figure 1A-C, Table I S1). Overall, 92% of donors 118 showed a specific CD4+ T cell response to at least one SARS-CoV-2 protein based on production 119 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprintthis version posted September 1, 2020. ; https://doi.org/10.1101/2020.08.27.20183319doi: medRxiv preprint 7 of at least one cytokine , where a positive response was defined as a 10% increase over control 120 stimulated samples (Table SI). No responses were detected to OC43 spike protein by ICC in the 121 same donors ( Figure S1B ), whereas 100% of donors produced cytokines in response to 122 PMA/ionomycin (Figure S1B). We did not observe cytokine production in CD8+ T cells from any 123 of the donors, albeit CD8+ T cells from all donors responded to influenza A/PR8/34 virus (PR8) 124 as well as to PMA/ionomycin (data not shown). This is likely because CD8+ T cells respond poorly 125 to whole protein antigens. Taken together, our data show that the vast majority of SARS -CoV-2 126 convalescent individuals have recall CD4+ responses to SARS -CoV-2 S or N proteins at 4 -10 127 weeks after initial symptoms, with IL-2 and TNF-a predominating over IFN-g. 128 129 Comparison of the CD4+ T cell response to S ARS-CoV-2 versus Influenza A virus by 130 multiparameter flow cytometry 131 As most adults are expected to have memory T cells specific for seasonal influenza virus, we 132 compared recall responses to influenza A/PR8/34 H1N1 virus (PR8) for all donors. 92% of SARS-133 CoV-2 convalescent patient samples showed strong CD4+ recall responses to PR8 stimulation 134 based on IFN-g producing T cells and the frequency of these responding cells was substantially 135 higher than responses to S and N proteins (Figure 1A). This was not due to insufficient S protein, 136 as increasing the dose from 1 to 5 µg per ml did not increase the frequency of responses (Figure 137 S2A). We also obtained similar responses using trivalent inactivated seasonal influenza vaccine 138 (TIV), which contains only influenza proteins (Figure S2B). Thus, the weaker response to SARS-139 CoV-2 S protein compared to influenza proteins is unlikely due to the use of live influenza virus 140 versus recombinant SARS-CoV-2 proteins, albeit it could be impacted by an incomplete set of 141 SARS-CoV-2 epitopes covered by including only 2 of the SARS-CoV-2 proteins. Some human T 142 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprintthis version posted September 1, 2020. ; https://doi.org/10.1101/2020.08.27.20183319doi: medRxiv preprint 8 cell studies use costimulation with anti-CD49d and anti -CD28 to increase the sensitivity of 143 detection with ICC (32), however we found no difference in the frequency of response to S with 144 or without additional costimulation (Figure S2C). We also repeated the assays 3 times for 2 of the 145 donors and obtained a similar frequency of responding T cells each time (Figure S2D). In contrast 146 to the results wi th SARS-CoV-2 convalescent patients, PBMC collected from healthy donors in 147 early March 2020 did not show detectable T cell responses to S, N or OC43 S whole proteins, 148 albeit all the healthy donors responded to PR8 (Figure S3). As will be discussed below, the lack 149 of detection of cross-reactivity in healthy donors may reflect the relative insensitivity of the ICC 150 assay compared to other methods of detecting cross-reactivity. 151 152 Analysis of T cell production of multiple cytokines showed that 80% of S -specific CD4+ T cells 153 produced only 1 cytokine . Influenza-specific CD4+ T cell responses were more multifunctional, 154 with 8.7% of PR8-specific CD4+ T cells as compared to 3.4% of S-specific T cells producing all 155 3 cytokines ( Figure 2A). We also noted that the ratio of IFN-g to TNF-α producing cells was 156 significantly higher among the PR8-specific CD4+ T cells than the S-specific CD4+ T cells 157 (Figure 2B) and this was independent of disease severity. 158 159 Analysis of CD27 and CD45RA expression on the S-specific and PR8-specific TNF-α-producing 160 CD4+ T cells indicated that the responding T cells were predominantly central memory T cells 161 (Figure 2C ). The activation markers HLA -DR and 4 -1BB are frequently used to determine 162 specific recall responses. Based on these markers, 100% of donors responded to influenza PR8, 163 whereas 69% responded to Spike and 85% to N. Examination of HLA- DR/4-1BB double positive 164 cells for cytokine production showed some discordance between activation markers and cytokine 165 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprintthis version posted September 1, 2020. ; https://doi.org/10.1101/2020.08.27.20183319doi: medRxiv preprint 9 producing cells (Figure 2D), with neither approach identifying 100% of the responding CD4+ T 166 cells. 167 168 Although there appeared to be a trend towards higher respon ses in donors with severe illness in 169 the first 4 weeks, using either activation markers and/or production of cytokines as a measure of 170 response, differences in ICC response based on disease severity were not significant (Figure 2E). 171 Taken together, our dat a show that SARS-CoV-2-specific T cells from patients in the early 172 convalescent phase are largely of the Tcm phenotype and respond to S and N proteins with a higher 173 ratio of TNF-a:IFN-g producing cells compared to the response to influenza virus, which shows a 174 more typical anti-viral IFN-g dominant response. 175 176 Recall responses of SARS-CoV-2 convalescent PBMC based on cytokine secretion 177 To further analyze cytokine production during recall responses to SARS -CoV-2 we collected 178 supernatants from ex vivo PBMC 48hrs post-stimulation with S, N or influenza PR8 by multiplex 179 bead array analysis of 13 cytokines (Figures 3, S4). 92% of SARS-CoV-2 convalescent donors 180 showed IFN-g production in response to SARS-CoV-2 S, whereas 100% showed IFN-g responses 181 to SARS-CoV-2 N and PR8, albeit the median level of IFN-γ in response PR8 was higher than that 182 observed in response to SARS-CoV-2 N or S (Figure 3A, S4A). 100% of patient PBMC produced 183 specific TNF-α responses in response to N, whereas 50% produced TNF -α in response to Spike 184 and 92% in response to PR8. N-specific responses showed substantially higher TNF-α responses 185 than S or PR8-specific responses (Figure 3B, S4A). IL-2 was produced in response to S or PR8, 186 but not in response to N -stimulation, whereas IL-10 was produced in all cases, albeit the highest 187 amount of IL -10 was observed in the N -stimulated cult ures (Figure 3 C, D , S4A ). IL -13 was 188 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprintthis version posted September 1, 2020. ; https://doi.org/10.1101/2020.08.27.20183319doi: medRxiv preprint 10 produced in response to S and PR8, but not N, whereas IL -6 was only observed with S and N 189 restimulation and not with PR8 (Figure 3 E, F). Similar to our findings with ICC, the ratio of IFN-190 g to TNF-α or IL-10 was highest in cultures stimulated with Influenza A virus (Figure 3G). We 191 did not detect IL-4, 5 or 17F in any of the cultures (data not shown). IL-9 and 17A were detected 192 in some cultures but did not show consistent increases with S or N stimulation, whereas IL-22 was 193 produced in response to S stimulation for some donors ( Figure S4 B). Healthy donor PBMC 194 produced IL-6, 10, IFN-g and TNF-α in response to N but not S proteins (Figure S4C). Stimulation 195 of SARS -CoV-2 convalescent PBMC as well as healthy donor PBMC with OC43 S protein 196 resulted in induction of IFN-g, TNF-a, IL-10 and IL-6, but not IL-2 (Figure S5A, B). 197 198 Overall, the multiplex cytokine assays show a predominant Th1 profile based on restimulation 199 with SARS-CoV-2 or seasonal hCoV-OC43 S protein, as well as reactivity of healthy donor PBMC 200 to SARS-CoV-2 N and OC43 S protein. SARS-CoV-2 convalescent patients’ PBMC showed a 201 lower IFN -γ to TNF -α or IFN -γ to IL -10 in response to SARS-CoV-2 proteins compared to 202 influenza virus restimulation. Particularly striking was the consistent and high-level production of 203 IL-10 in response to N in all SARS-CoV-2 convalescent PBMC tested. 204 205 Peripheral T follicular helper as well as T effector re sponses to SARS-CoV-2 antigens correlate 206 with serum antibodies and neutralization titers 207 T follicular helper (Tfh) responses are important for the generation of long -lived antibody 208 responses (33). Although fully differentiated Tfh are normally found in the lymphoid organs, their 209 peripheral blood precursors, pTfh, can be detected in the blood based on expression of CCR7 and 210 CXCR5 (34, 35). Here we used expression of IL -2 by ICC combined with CCR7 and CXCR5 211 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprintthis version posted September 1, 2020. ; https://doi.org/10.1101/2020.08.27.20183319doi: medRxiv preprint 11 expression to detect antigen -specific pTfh cells in SARS -CoV-2 convalescent patient PBMC 212 following restimulation with S, N or PR8 ( Figure 4A). 46% of samples showed S -specific IL-2 213 producing pTfh, 54% N-specific and 100% of PBMC samples showed PR8 -specific pTfh. For 214 samples collected during the first 4 weeks post -symptoms, pTfh responses to S were higher in 215 severe compare to mild cases (Figure 4A, B). We also compared pTfh responses of PBMC from 216 the SARS-CoV-2 convalescent patients with IgG and IgA responses to N and RBD based on serum 217 ELISA (36) and neutralization data (Figure 4C,D). There was a positive correlation between the 218 frequency of IL-2+pTfh and N-specific IgG (R=0.57, p<0.05). The correlation between IL-2+pTfh 219 and RBD -specific IgG showed a similar positive trend (R=0. 29) but did not reach statistical 220 significance. Similarly, no significant correlation with S -specific IgG was observed (data not 221 shown). There was a significant positive correlation between the pTfh response to RBD -specific 222 IgA (R=0.6 6, p< 0.05). There was also a strong correlation between WT SARS -CoV-2 IC50 223 neutralization (modified PRNT assay) by patient sera and the pTfh response (R=0.8 4, p<0.001). 224 Similar correlations were obtained by calculating the area under the curve (AUC) in a surrogate 225 neutralization ELISA with patient sera, human ACE2 and immobilized S-RBD (R=-0.75, p<0.01) 226 (Figure 4E). 227 228 Similarly, there was a positive correlation between the frequency of IL-2+ CD4+ T cells and N-229 specific IgG (R=0.61, p<0.05), and a positive trend between IL-2+ CD4+ T cells and RBD-specific 230 IgG (R=0.36), albeit not significant ( Figure 5A). There was also a positive correlation between 231 IL-2+ CD4+ and RBD-specific IgA (R=0.60, p<0.05) (Figure 5B). A strong positive correlation 232 was also found between virus neutralization and IL-2+ CD4+ T cells (R=0.82, p<0. 001), and 233 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprintthis version posted September 1, 2020. ; https://doi.org/10.1101/2020.08.27.20183319doi: medRxiv preprint 12 between S -RBD IgG AUC and IL-2+ CD4+ T cells. (Figure 5C). There was also a positive 234 correlation between virus neutralization and disease severity (Figure 5D). 235 236 Thus, pTfh responses can be detected in 6 out of 13 SARS-CoV-2 convalescent PBMC responding 237 to S protein , 7/13 in response to N whereas 13/13 showed a Tfh response to influenza A virus. 238 Both pTfh and T effector responses correlated strongly with the neutralization titers observed in 239 the same donors, with the highest neutralization activity correlating with disease severity. 240 241 CD4+ and CD8+ T cell proliferative responses to peptide pools 242 Much of the published work on SARS-CoV-2 specific T cells has focused on peptide pools and 243 these are more effective in inducing CD8+ recall responses than intact proteins. Therefore, we 244 used peptide pools encompassing the RBD, transmembrane (TM) and cytoplasmic regions of S as 245 well as N, Envelope (E) and Matrix (M) to stimulate PBMC from the same patient samples used 246 for ICC. As T cell proliferation to virus antigens has previously been associated with ability to 247 control the virus (37, 38), we labelled PBMC CFSE and assessed the proliferation of the T cells 248 after 7 days by flow cytometry in response to each peptide pool (Figure 6A, with gating strategy 249 shown in Figure S6). Proliferation responses of total T cells to at least one antigen was observed 250 in 12/13 donors (Table S2). There was considerable variability between donors. Generally, donors 251 who made strong proliferative responses had strong responses to all antigens tested. However, 252 antigen-specific proliferation did not correlate significantly with disease severity (Figure 6B, C, 253 D). We also examined PBMC from a SARS-CoV-1 patient taken 17 years post-illness and 254 observed modest reactivity to the N peptide pool (Figure 6B, C). We next broke down responses 255 into CD4+ and CD8+ T cell responses for each peptide pool. For most subjects and antigens, CD4+ 256 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprintthis version posted September 1, 2020. ; https://doi.org/10.1101/2020.08.27.20183319doi: medRxiv preprint 13 T cell proliferative responses were substantially higher than CD8+ T cell responses, independent 257 of disease severity (Figure 6E). 258 259 Although CD8+ cytotoxic T cells are classically associated with virus-infected cell killing, CD4+ 260 granzyme+ cytotoxic T cells can be a significant part of the human anti-viral T cell responses (37, 261 39-41). Therefore, we also assessed IFN -g and granzyme B levels by flow cytometry in the 262 CFSElow responding CD4+ and CD8+ T cells (Figure 7A, B). Of note, the samples from subjects 263 with severe and moderate disease tended to have a higher proportion of CD4+ IFN-g/granzyme B 264 co-producing T cells than mild, however this was not universally the case, as we also saw a high 265 proportion of IFN-g/granzyme B expressing T cells in the asymptomatic donor. 266 267 The frequency of proliferating IFN-g/granzyme B co-producing T cells in response to S peptide 268 pools correlated with the frequency of IL-2 producing pTfh in response to intact Spike (Figure 7C) 269 as well as with virus neutralization titers (Figure 7D). In addition, proliferating IFN-g/granzyme B 270 producing cells in response to E or M peptide pools correlated with serum neutralization titers 271 (Figure 7 E, F). Thus, a strong CD4+ response overall, whether based on analysis of whole protein 272 or peptide stimulation, correlates with strong neutralization responses. 273 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprintthis version posted September 1, 2020. ; https://doi.org/10.1101/2020.08.27.20183319doi: medRxiv preprint 14

274 In this study, we have conducted a systematic examination of T cell recall responses in PBMC 275 taken in the early convalescent phase (4-12 weeks post-symptoms) of COVID-19 in response to 276 SARS-CoV-2 recombinant proteins as well as to peptide arrays . The use of recombinant proteins 277 is relevant because it allows us to assess the response in the context of antigen presentation, and 278 the use of the fully glycosylated Spike trimer is important in mimicking the form of antigen that is 279 presented by intact virus. A T cell response was detected in all SARS-CoV-2 convalescent patients, 280 with 92% responding based on ICC responses to recombinant SARS-CoV-2 proteins and 100% 281 responding based on multiplex cytokine assays . CD8+ T cell respon ses were not detected in 282 response to whole protein restimulation by ICC but were identified in 12 out of 13 patients based 283 on proliferation in response to restimulation with peptide pools encompassing N, E, M or S, albeit 284 with varying frequencies. While several other studies have identified Th1 responses in response to 285 SARS-CoV-2 peptide or protein stimulation (11, 23, 25, 27, 29) , our comparison of recall 286 responses to influenza within the same donors highlights some key differences in SARS -CoV-2 287 versus influenza-specific T cell responses. 288 289 Intracellular cytokine staining revealed that SARS-CoV-2-specific CD4+ recall responses exhibit 290 a hierarchy of IL-2>TNF-a>IFN-g, whereas influenza A virus-specific T cells show IFN-g>IL-2> 291 TNF-a based on frequency of cytokine producing cells. This altered Th1 profile in SARS-CoV-2-292 specific T cells could contribute to increased inflammation with poorer viral control compared to 293 influenza virus-specific T cells. It was possible that these differences are due to the use of whole 294 influenza A virus compared to r ecombinant proteins . However, control experiments showed 295 indistinguishable frequencies of responding CD4+ T cells producing IFN -g, TNF -a and IL -2 296 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprintthis version posted September 1, 2020. ; https://doi.org/10.1101/2020.08.27.20183319doi: medRxiv preprint 15 whether samples were restimulated with live influenza virus or with TIV, the inactivated influenza 297 vaccine which is dominated by the hema gglutinin protein. It is unlikely that the altered cytokine 298 response in response to SARS -CoV-2 antigens is driven by disease severity, as mild and severe 299 patients were distributed throughout the plots showing this altered ratio (Fig. 2B, 3G). On the other 300 hand, the lower frequency of multifunctional cells in the SARS -CoV-2 specific as compared to 301 influenza A-specific CD4+ recall responses seems to be heavily weighted by the severe cases (Fig. 302 2B) and could reflect COVID-19-specific exhaustion, as has been suggested by other studies based 303 on activation/exhaustion markers (42, 43). The influenza-specific recall responses we observed are 304 likely due to a lifetime of seasonal exposure and/or vaccination, whereas the SARS -CoV-2- 305 specific responses are more recent. However, it is unlikely that the time since exposure is driving 306 the altered cytokine profile we observe , as the cytokine profile observed in recall responses 307 generally reflects the epigenetic profile imprinted during priming (44). 308 309 Multiplex analysis of cytokines in supernatants of PBMC following SARS -CoV-2 antigen 310 stimulation also revealed lower IFN-g to TNF-a ratios of SARS -CoV-2 compared to influenza -311 specific responses as well as higher levels of IL-10 and IL-6. The ICC flow cytometry assay allows 312 one to clearly identify the source of the cytokines as CD4+ T cells, whereas the multiplex cytokine 313 assay reflects total amount of secreted cytokine and can also reflect cytokines secreted from other 314 cells, such as monocytes or NK cells, in response to the activated T cells. IL-10 can be produced 315 by both T cells and antigen presentin g cells, whereas IL -6 is likely coming from monocytes 316 responding to the activated T cells in the PBMC culture. Particularly striking in our study was the 317 high level of IL-10 detected in the supernatants of N -stimulated cultures, which could contribute 318 to impaired antigen presentation and immunosuppression (45). Further investigation is required to 319 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprintthis version posted September 1, 2020. ; https://doi.org/10.1101/2020.08.27.20183319doi: medRxiv preprint 16 determine whether N-specific responses are immunosuppressive , which would have significant 320 implications for vaccine design. 321 322 A potential caveat to our findings is that we included only 2 of the SARS-CoV-2 proteins in our 323 cytokine analysis and not the full spectrum of SARS -CoV-2 antigens. However, the cytokine 324 profile we observed in the supernatants of S and N stimulated PBMC is quite similar to that 325 reported by Weiskopf et al for SARS -CoV-2 ARDS patient PBMC collected 3 weeks after ICU 326 admission and stimulated with peptide megapeptide pools covering most of the SARS -CoV-2 327 proteins (23). There were, however, some differences noted, such as their detection of IL -17A, 328 which we did not detect in our assays. 329 330 The cytokine profile we detect in the supernatants of SARS -CoV-2 convalescent PBMC after 331 antigen stimulation is similar to the overall cytokine profile reported at the acute phase of infection, 332 including high levels of IL-6, IL-10 and TNF-a (46, 47). This is consistent with the evidence that 333 memory T cells are imprinted by the acute inflammatory milieu (44). Schultheiss et al.(48) recently 334 analyzed total PBMC from SARS -CoV-2 active and early convalescent patients and also noted 335 that total CD4+ T cells showed an altered non-classical Th1 profile, similar to what we observe 336 here with antigen -specific T cell responses. They also noted Th17 responses, which were not 337 consistently observed in the antigen-specific T cells in our cohort. 338 339 Of note, we observed a disconnect between ICC responses and analysis of T cell responses to S 340 and N based on activation markers. This was not unique to SARS-CoV-2, however, as we observed 341 a similar disconnect with influenza A -specific T cell cytokine response and activation markers 342 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprintthis version posted September 1, 2020. ; https://doi.org/10.1101/2020.08.27.20183319doi: medRxiv preprint 17 (data not shown). It is possible that some antigen-specific T cells are not making cytokines in the 343 time frame analyzed or that some of these activation markers are induced on memory T cells by 344 bystander effects (cytokines). We suggest, that functional readouts based on cytokines may be 345 more relevant to understanding protective immunity to SARS -CoV-2 than use of activation 346 markers. 347 348 Peripheral T follicular helper responses (IL-2+CCR7+CXCR5+) were detected in 62% of PBMC 349 after S or N stimulation, and strongly correlated with virus neutralization activity of sera based on 350 neutralization of SARS-CoV-2 as well as a surrogate neutralization ELISA for binding to RBD. 351 The strong correlation with IgA might reflect the recently reported role for IgA in SARS-CoV-2 352 neutralization (49). Our findings are similar to those of Ni et al. (11) who showed a correlation 353 between total N-specific T cells measured by EL ISpot with neutralizing antibody titers. Of note, 354 total T effector responses to N and S as well as IFN -g or IFN-g and granzyme B r esponses to E 355 and M peptide pools also correlated with virus neutralization, suggesting that a strong CD4+ T cell 356 response in general correlated with effective virus neutralization , whether we used peptide pools 357 or intact antigens for these recall assays. Of note, 100% of donors showed pTfh responses to 358 influenza virus and the response was generally of higher frequency . Thus, the Tfh respons e to 359 SARS-CoV-2 in convalescent subjects is weaker than that observed in response to influenza virus 360 restimulation. 361 362 Several recent studies have revealed responses of healthy donors to SARS -CoV-2 peptides (21-363 23, 28, 31). It has also been suggested that prior exposure to seasonal coronaviruses might allow 364 some cross-protective immunity to SARS -CoV-2 (30). Our ICC assays did not reveal responses 365 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprintthis version posted September 1, 2020. ; https://doi.org/10.1101/2020.08.27.20183319doi: medRxiv preprint 18 of SARS -CoV-2 convalescent patients or healthy donors to seasonal OC43 spike protein . 366 However, such responses were detected in supernatants based on the cytokines IFN-g, TNF, IL-6 367 and IL -10, but not IL -2. This may reflect the lower sensitivity of the overnight ICC assay, 368 compared to assessme nt of cyto kines in the supernatant at 48hrs. Healthy donors similarly 369 responded to OC43 spike but with approximately 10 -fold weaker responses than SARS-CoV-2 370 convalescent patients suggesting that recent boosting with SARS -CoV-2 might enhance such 371 responses. Healthy donors also responded to SARS-CoV-2 N but not S based on release of IFN-g, 372 TNF-a and IL-6 in the supernatant. We also detected proliferative responses to the N peptide pool 373 of a SARS-CoV1 patient, 17 years post-illness, similar to results recently reported (21). 374 375 Proliferative responses to SARS-CoV-2 peptide pools showed that CD4+ responses predominated 376 over CD8+ T cell responses, which might contribute to the pathophysiology of COVID19. On the 377 other hand, many of these CD4+ T cells co-produced IFN-g and granzyme B, suggesting cytotoxic 378 potential. As airway epithelial cells, the target of SARS-CoV-2 infection, can express MHC II (50-379 52), these granzyme B positive cells may be relevant to viral control. 380 381 A recent study of early T cell responses to SARS -CoV-2 showed delayed T cell responses, 382 compared to antibody responses in the first two weeks post symptom onset, but with T cell 383 responses increasing at >3 weeks (53). Although we did not do a kinetic analysis, the data on 384 convalescent samples collected at 4-12 weeks post-symptoms, are consistent with a peak response 385 around 4 weeks, and falling off thereafter. Th ese kinetics are similar to what was observed in the 386 recall response to the 2009 influenza virus pandemic, where peripheral blood CD4+ and CD8+ T 387 cell responses to whole H1N1 restimulation peaked at about 3 -4 weeks post symptoms and then 388 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprintthis version posted September 1, 2020. ; https://doi.org/10.1101/2020.08.27.20183319doi: medRxiv preprint 19 fell off gradually (54). We did not see a consistent difference between severe and mild cases in 389 terms of magnitude of the T cell response, albeit this may be limited by sample size. 390 In sum, our study shows robust T cell recall responses in SARS -CoV-2 convalescent subjects at 391 4-12 weeks post-symptoms. Based on proliferation, ICC or multiplex ELISA, all donors showed 392 SARS-CoV-2-specific T cell responses . By 4 weeks post-SARS-CoV-2 infection, most s ubjects 393 exhibit a strong CD4+ Th1 recall response, with a less predominant CD8+ T cell response and an 394 altered cytokine profile with more TNF-a and less IFN-g compared to responses to influenza virus 395 in the same donors. In addition, pTfh responses to SARS-CoV-2 were weaker than that to influenza 396 A virus. SARS-CoV-2 N-specific T cell responses were associated with strong induction of IL-10, 397 suggesting that N protein might contribute to immunosuppression. This could have important 398 implications for vaccine design. Taken together, these results suggest that CD4+ T cell responses 399 are more inflammatory than influenza-specific recall responses and show a weaker Tfh response, 400 potentially contributing to disease. The strong correlation between the N- or S -specific pTfh 401 response or the IFN -g/granzyme B+ proliferative response and neutralization capacity suggests 402 that these responses should be incorporated into vaccine design and testing. 403 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprintthis version posted September 1, 2020. ; https://doi.org/10.1101/2020.08.27.20183319doi: medRxiv preprint 20

404 Human subjects and study approval 405 Written informed consent was obtained from COVID -19 convalescent and healthy blood donors 406 before leukapheresis or peripheral blood samples were obtained. Individuals with recovered 407 COVID-19 infection that was confirmed by positive nasopharyngeal COVID -19 P CR upon 408 presentation, were leukapheresed after resolution of symptoms through an REB approved protocol 409 (St. Michael's Hospital REB20-044c to MO . Additional healthy donors were recruited at the 410 University of Toronto ( REB# 00027673 to THW) . All human subject s research was done in 411 compliance with the Declaration of Helsinki. 412 413 Human PBMC isolation 414 PBMCs were isolated from whole blood of healthy human donors by density centrifugation using 415 Ficoll-Paque PLUS (GE Healthcare). PBMCs were cryopreserved in 10% DMSO in AIM-V media 416 (Gibco) before use. 417 418 Virus and viral antigens 419 The human codon-optimized cDNA encoding the OC43 spike protein (AAT84354.1) was 420 synthesized by GeneArt. The soluble OC43 spike construct includes residues 15-1295, followed 421 by a T4 fibritin trimerization motif, a TEV cleavage site, and a 6xHis-tag. The 20 amino acid 422 human cystatin secretion signal was added N-terminal to the spike sequence. To stabilize the pre-423 fusion state of the OC43 spike trimer, residues 1070-1071 (AL) were mutated to two proline 424 residues (PP) as described for other spike proteins (55). The human codon-optimized cDNA 425 encoding the SARS-CoV-2 spike protein (YP_009724390) was synthesized by GenScript. The 426 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprintthis version posted September 1, 2020. ; https://doi.org/10.1101/2020.08.27.20183319doi: medRxiv preprint 21 soluble spike trimer construct includes residues 1 -1211, followed by a T4 fibritin trimerization 427 motif, a 6xHis -tag and an AviTag biotinylation motif (56). Residues 682 –685 (RRAR) were 428 mutated to SSAS to remove the fur in cleavage site on the SARS -CoV-2 spike protein. Residues 429 986–987 (KV) were mutated to two proline residues (PP) to stabilize the pre-fusion form. 430 431 The spike proteins were cloned into a piggyBac -based inducible expression vector PB -T-PAF. 432 Inducible stable cell lines were generated in Freestyle 293 -F cells (Thermofisher) as previously 433 described (36, 57). For the OC43 spike protein, the stable cells were grown as an adherent culture 434 in DMEM/F12 medium supplemented with 3% (v/v) FBS. For the SARS-CoV-2 spike protein, the 435 stable cells were grown in suspension culture in Freestyle 293 expression medium (Thermofisher). 436 Protein expression was induced by the addition of 1 µg/mL doxycycline. The secreted proteins 437 were purified from the tissue culture medium using Ni -NTA resin. The proteins were further 438 purified by size-exclusion chromatography using a Superose 6 Increase column (GE healthcare). 439 The quality of the purified spike protein trimers was assessed using negative stain electron 440 microscopy. 441 442 Nucleocapsid1-419 (N) expressed as a N-terminally tagged HIS-GST-TEV fusion was purified from 443 bacteria and kindly provided by Frank Sicheri, Mt. Sinai Hospital, as described in (17). 444 445 Endotoxin levels were measured in S and N proteins using the Toxin Sensor Chromogenic LAL 446 Endotoxin Assay Kit, from GenScript/VWR, Cat # L00350C. Final concentrations of LPS was <1 447 EU per well (0.18 for S, 0.43 for N). Influenza virus strain A/Peurto Rico/8/1934 (PR8) was grown 448 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprintthis version posted September 1, 2020. ; https://doi.org/10.1101/2020.08.27.20183319doi: medRxiv preprint 22 in embryonated chicken eggs and tissue culture infectious dose determined by infection of MDCK 449 cells (58). 450 451 15-mer peptides overlap ped by 11 amino acids spanning most of full protein sequence of N, 452 membrane (M), envelope (Env), and RBD/TM/cytoplasmic domains of S protein of SARS-CoV-453 2 were synthesized (GeneScript). To stimulate PBMC, N-master peptide pool with102 peptides, 454 Env-master peptide pool with 12 peptides, M-master peptide pool with 49 peptides and S -master 455 peptide pool with 49 peptides were used in the study. 456 457 T cell stimulation assay 458 For all stimulation assays, cryopreserved PBMCs were thawed at 37°C, washed twice with PBS 459 and cultured in complete media (RPMI 1640 supplemented with 10% FBS, 2 -ME, sodium 460 pyruvate, penicillin, streptomycin and non -essential amino acids (Gibco)) at 37°C with 5% CO 2. 461 2x106 PBMCs were plated per well in 96-well round bottom plates for 18h with 1 μg/ml S, 1 μg/ml 462 N, 3 μg/ml OC43 S or 100 HAU/ml live PR8. PBMCs were cultured with 1 μg/ml BSA (Sigma-463 Aldrich) as a negative control. GolgiStop (BD Biosciences) containing monensin and GolgiPlug 464 (BD Biosciences) containing brefeldin A was added in the last 6h of the culture. As a positive 465 control, 50 ng/ml PMA (Sigma -Aldrich), 1 μg/ml ionomycin (Sigma -Aldrich), GolgiStop and 466 GolgiPlug were added to PBMCs cultured with complete media in the last 6h of culture. 467 468 To assess T cell recall responses to live PR8 c ompared to TIV (FLUZONE® High -Dose), cells 469 were either cultured with complete media, 100 HAU/ml PR8 or 1 μg/ml TIV for 18h. To determine 470 whether the addition of agonistic co -stimulatory antibodies increased the sensitivity of detection 471 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprintthis version posted September 1, 2020. ; https://doi.org/10.1101/2020.08.27.20183319doi: medRxiv preprint 23 of ICC by flow cytometry, PBMCs were stimulated with 1ug/ml S or 1 μg/ml BSA, either with or 472 without 2 μg/ml anti -CD28 and 2 μg/ml anti -CD49d (BD Biosciences) for 18h. GolgiStop and 473 GolgiPlug were added in the last 6h of these cultures. 474 475 Intracellular cytokine staining 476 After culture, PBMCs were washed with PBS containing 2% FBS (FACS buffer). Cells were first 477 stained with anti -human CCR7 at 37°C for 10 min, followed by staining with Fixable Viability 478 Dye eFluor™ 506 (eBiosciences) to discern viable cells, and with anti-human CD3, CD4, CD27, 479 CD45RA, CXCR5, 4-1BB and HLA-DR for 20 min at 4°C. Cells were washed twice with FACS 480 buffer, then fixed with BD Cytofix/Cytoperm buffer (BD Biosciences) for 20 min. Following 481 fixation and permeabilization, cells were washed twice with 1X B D Perm/Wash buffer (BD 482 Biosciences) and stained with anti -human IFN-γ, TNF-α, IL-2 and IL -17A for 15 min at 4°C. 483 Antibodies used are as listed in Table S3. Samples were washed twice, then resuspended in FACS 484 buffer and acquired on the BD LSRFortessa X-20 flow cytometer using FACSDiva software. 485 486 Multiplex cytokine bead assay 487 2x106 PBMCs were seeded per well in 96-well round bottom plates with 1 μg/ml S, 1 μg/ml N, 3 488 μg/ml OC43 S, 1 μg/ml BSA or 100 HAU/ml PR8. Cell culture supernatants were collected after 489 48h of incubation. Cytokines in the supernatants were measured using the Human Th Cytokine 490 Panel (12-plex) LEGENDplex kit (Biolegend) with capture reagents specific for IL-2, IL-4, IL-5, 491 IL-6, IL-9, IL-10, IL-13, IL-17A, IL-17F, IL-22, IFN-γ and TNF-α. The assay was performed as 492 per the manufacturer’s instructions using a V -bottom plate. Samples were acquired on the BD 493 LSRFortessa X-20 flow cytometer. 494 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprintthis version posted September 1, 2020. ; https://doi.org/10.1101/2020.08.27.20183319doi: medRxiv preprint 24 495 CFSE T cell proliferation assay 496 PBMC (2 x 106 cells/ml) were pre-labeled with 5µM of carboxyfluorescein diacetate succinimidyl 497 ester (CFSE; Thermo Fisher Scientific) in PBS with 2.5% FBS for 8 minutes in 37 °C water bath. 498 Excessive CFSE dye was removed by using 100% FBS and further rinsed with R-10 [RPMI1640, 499 FBS, Pen/Strep (Thermo Fisher Scientific), Glutamax (Thermo Fisher Scientific) and sodium 500 pyruvate (Thermo Fisher Scientific)]. Cells were then resuspended in R -10 [supplemented with 501 IU-IL 2 (BioLegend) and 2-mercaptoethanol (Themor Fisher)] and plated at 0.4x 106 cells per well 502 in a 96-well round-bottomed polystyrene plate at a final volume of 200 µl. These cells were pre -503 stimulated with 0.1µg of S, E, N and M master peptide pools or DMSO (negative control) or SEB 504 (positive control) for 5 days. At day 6, cells were re -stimulated with 1 µg/ml of master peptide 505 pools and the exocytosis was blocked by the addition of BD GolgiStop and BD GolgiPlug for 506 another 24 hours. At day 7, cells were prepared for flow c ytometry staining. LIVE/DEAD™ 507 fixable blue dead cell stain (Thermo Fisher Scientific) was used to determine the viability of cells 508 and then pre-blocked with Fc receptor blocking solution (Human TruStain FcX TM; BioLegend) 509 prior to extracellular staining with anti-human CD3 [APC-Cy7: clone SK7 (BD)], anti-human CD4 510 [BV711: clone SK3 (BD)] and anti-human CD8 [PE: clone HIT8a (BD)]. Cells were then fixed 511 with BD Cytofix and permeabilized with BD Perm/Wash as per the manufacturer’s protocol and 512 stained with anti-human IFNg- [APC: clone 4S.B3 (BD)] and anti-human granzyme B [BV421: 513 clone GB11(BD)]. Samples were acquired on the BD LSRFortessa X -20 flow cytometer. Net 514 peptide pool induced CFSElow responses were calculated as the percentage of CFSElow cells after 515 stimulation with master pool peptides minus the percentage of CFSElow cells after stimulation with 516 DMSO. 517 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprintthis version posted September 1, 2020. ; https://doi.org/10.1101/2020.08.27.20183319doi: medRxiv preprint 25 518 WT SARS-CoV-2 Neutralization Assay 519 100 μl of Vero E6 cells were seeded into a 96 well plate at 0.3x106 cells/mL and were incubated 520 overnight for attachment. The following day, patient serum was heat inactivated at 56 °C for 30 521 minutes, then serially diluted 8 times, 2-fold downwards starting at 1:10. Equal volumes of SARS-522 CoV-2 were added to all wells with a final concentration of 100 TCID/well. The plate was 523 incubated for 1h, shaking every 15 minutes. After incubation, all the media from the Vero E6 cells 524 was removed and 50 μl of the SARS-CoV-2/Serum co-culture was used to inoculate the Vero E6 525 cells. The infection was done for 1h, shaking every 15 minutes. After infection, the inoculum was 526 removed and growth media was added. CPE was tracked over the course of 5 days. Samples were 527 run in quadruplicates. 528 529 Protein-based surrogate neutralization ELISA 530 A protein-based surrogate neutralization ELISA was performed as described in (36). Essentially, 531 100 ng of purified RBD express ed in FreeStyle 293 -F cells was immobilized overnight onto 96 -532 well Immulon HBX plates (2 µg/ml), blocked, and incubated with four, 2-fold dilutions of patient 533 samples, starting at 4 µl. Biotinylated ACE2 purified from FreeStyle 293 -F cells was added (50 534 ng/well, incubated for 1hr), followed by streptavidin poly-HRP (Sigma, #S2438; 22 ng). 1-Step™ 535 Ultra TMB-ELISA Substrate Solution (ThermoFisher, #34029) was added for 7.5 min at room 536 temperature and the reaction was quenched with 50 µL stop solution containi ng 0.16 N sulfuric 537 acid (ThermoFisher, #N600) and the optical density at 450 nm were read. The area under the curve 538 of each dilution series for each patient plasma sample was calculated in R (version 4.0.1). 539 540 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprintthis version posted September 1, 2020. ; https://doi.org/10.1101/2020.08.27.20183319doi: medRxiv preprint 26 Data and Statistical analysis 541 Flow cytometry da ta were analyzed using FlowJo v10. Multiplex cytokine bead data were 542 analyzed using the LEGENDplex Data Analysis Software v8. All statistical and graphical analyses 543 were performed using Graphpad Prism v6. Illustrations were created with Biorender.com. When 544 data are shown in the absence of the control group, the values are calculated by subtracting 545

signal as indicated by “Δ” in panel labels . Background signal is defined by the 546 frequency of cells expressing a particular cytokine, or concentration of an analyte, in wells cultured 547 with BSA. The response was considered positive if the response to SARS-CoV-2 antigen was 10% 548 higher than the response to BSA. For multiplex cytokine data, the limits of detection are indicated 549 with dashed lines. Pair-wise co mparisons were made by two -tailed Wilcoxon test, one -way 550 ANOVA with Holm -Sidak’s multiple comparisons test or nonparametric Dunn’s multiple 551 comparisons test as indicated in figure legends. Correlation analyses were performed by computing 552 the Pearson or Spearman correlation coefficient. Statistical outliers were excluded from analyses 553 by Grubb’s test, but all data points are displayed in figure panels. 554 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprintthis version posted September 1, 2020. ; https://doi.org/10.1101/2020.08.27.20183319doi: medRxiv preprint 27 Author contributions 555 JCL, WHK, PB, JL, KTA and BR performed experiments 556 FYY, JCL and MG processed patient samples 557 JL and JMR provided purified S protein, 558 AM, AC and MO recruited patients 559 SM provided SARS-CoV-2 virus 560 KTA, BR and AGC provided direct binding ELISA surrogate neutralization ELISA data 561 JCL, WHK, PB, MO and THW designed the experiments and wrote the paper. 562 563 Acknowledgments 564 We thank Birinder Ghumman for technical assistance, Payman Samavarchi-Tehrani, Derek 565 Ceccarelli and Frank Sicheri, Mt. Sinai Hospital Toronto, for recombinant N protein purification 566 and Jennifer Gommerman for helpful discussion. This research was funded by a FAST grant from 567 the Thistledown foundation (to T.H.W.) and by a grant VR1-172711 from the Canadian Institutes 568 of Health research to T.H.W., M.O. and A.C.G. M.O. receives funding from the Ontario HIV 569 Treatment Network (OHTN), the Li Ka Shing Knowledge Institute, and the Juan and Stefania fund 570 for COVID-19 and other virus infections. Funding for the development of the assays in the Gingras 571 lab was provided through generous donations from the Royal Bank of Canada (RBC), QuestCap 572 and the Krembil Foundation to the Sinai Health System Foundation; the equipment used is housed 573 in the Network Biology Collaborative Centre at the Lunenfeld-Tanenbaum Research Institute, a 574 facility supported by Canada Foundation for Innovation funding, by the Ontarian Government and 575 by Genome Canada and Ontario Genomics (OGI-139). JCL and KTA were recipients of Ontario 576 Graduate Scholarships. 577 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprintthis version posted September 1, 2020. ; https://doi.org/10.1101/2020.08.27.20183319doi: medRxiv preprint 28

578 579 1. Zhu N, Zhang D, Wang W, Li X, Yang B, Song J, et al. A Novel Coronavirus from 580 Patients with Pneumonia in China, 2019. N Engl J Med. 2020;382(8):727-33. 581 2. Zhou P, Yang XL, Wang XG, Hu B, Zhang L, Zhang W, et al. A pneumonia outbreak 582 associated with a new coronavirus of probable bat origin. Nature. 2020;579(7798):270-3. 583 3. Wu F, Zhao S, Yu B, Chen YM, Wang W, Song ZG, et al. A new coronavirus associated 584 with human respiratory disease in China. Nature. 2020;579(7798):265-9. 585 4. Chandrashekar A, Liu J, Martinot AJ, McMahan K, Mercado NB, Peter L, et al. SARS-586 CoV-2 infection protects against rechallenge in rhesus macaques. Science 587 101126/scienceabc4776. 2020. 588 5. Bao L, Deng W, Gao H, Xiao C, Liu J, Xue J, et al. Lack of Reinfection in Rhesus 589 Macaques Infected with SARS-CoV-2. BioRXIV 590 https://doiorg/101101/20200313990226. 2020. 591 6. Long QX, Liu BZ, Deng HJ, Wu GC, Deng K, Chen YK, et al. Antibody responses to 592 SARS-CoV-2 in patients with COVID-19. Nat Med. 2020;26(6):845-8. 593 7. Long QX, Tang XJ, Shi QL, Li Q, Deng HJ, Yuan J, et al. Clinical and immunological 594 assessment of asymptomatic SARS-CoV-2 infections. Nat Med 101038/s41591-020-595 0965-6. 2020. 596 8. Zhao J, Yuan Q, Wang H, Liu W, Liao X, Su Y, et al. Antibody responses to SARS-597 CoV-2 in patients of novel coronavirus disease 2019. Clin Infect Dis 101093/cid/ciaa344. 598 2020. 599 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprintthis version posted September 1, 2020. ; https://doi.org/10.1101/2020.08.27.20183319doi: medRxiv preprint 29 9. Okba NMA, Muller MA, Li W, Wang C, GeurtsvanKessel CH, Corman VM, et al. 600 Severe Acute Respiratory Syndrome Coronavirus 2-Specific Antibody Responses in 601 Coronavirus Disease Patients. Emerg Infect Dis. 2020;26(7):1478-88. 602 10. Thevarajan I, Nguyen THO, Koutsakos M, Druce J, Caly L, van de Sandt CE, et al. 603 Breadth of concomitant immune responses prior to patient recovery: a case report of non-604 severe COVID-19. Nat Med. 2020;26(4):453-5. 605 11. Ni L, Ye F, Cheng ML, Feng Y, Deng YQ, Zhao H, et al. Detection of SARS-CoV-2-606 Specific Humoral and Cellular Immunity in COVID-19 Convalescent Individuals. 607 Immunity. 2020;52(6):971-7 e3. 608 12. Crawford KH, Dingens AS, Eguia R, Wolf CR, Wilcox N, and al. e. Dynamics of 609 neutralizing antibody titers in the months after SARS-CoV-2 infection. MedRXIv 610 2020;doi: https://doi.org/10.1101/2020.08.06.20169367. 611 13. Wajnberg A, Amanat F, Firpo A, Altman D, Bailey M, and al. e. SARS-CoV-2 infection 612 induces robust, neutralizing antibody responses that are stable for at least three months. 613 MedRXiv doi: https://doiorg/101101/2020071420151126. 2020. 614 14. Ibarrondo FJ, Fulcher JA, Goodman-Meza D, Elliott J, Hofmann C, Hausner MA, et al. 615 Rapid Decay of Anti-SARS-CoV-2 Antibodies in Persons with Mild Covid-19. N Engl J 616 Med 101056/NEJMc2025179. 2020. 617 15. Seow J, Graham C, Merrick B, Acors S, Steel KAJ, Hemmings O, et al. Longitudinal 618 evaluation and decline of antibody responses in SARS-CoV-2 infection. MedRIXIV 619 https://wwwmedrxivorg/content/101101/2020070920148429v1 620 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprintthis version posted September 1, 2020. ; https://doi.org/10.1101/2020.08.27.20183319doi: medRxiv preprint 30 16. Rodda LB, Netland J, Shehata L, Pruner KB, Morawski P, and al. e. Functional SARS-621 CoV-2-specific immune memory persists after mild COVID-19. MedRXIv doi: 622 https://doiorg/101101/2020081120171843. 2020 623 17. Isho B, Abe KT, Zuo M, Jamal AJ, Rathod B, and al. e. Evidence for sustained mucosal 624 and systemic antibody responses to SARS-CoV-2 antigens in COVID-19 patients. 625 medRxiv 2020080120166553; doi: https://doiorg/101101/2020080120166553. 626 18. Crotty S. T follicular helper cell differentiation, function, and roles in disease. Immunity. 627 2014;41(4):529-42. 628 19. Tang F, Quan Y, Xin ZT, Wrammert J, Ma MJ, Lv H, et al. Lack of peripheral memory B 629 cell responses in recovered patients with severe acute respiratory syndrome: a six-year 630 follow-up study. J Immunol. 2011;186(12):7264-8. 631 20. Ng OW, Chia A, Tan AT, Jadi RS, Leong HN, Bertoletti A, et al. Memory T cell 632 responses targeting the SARS coronavirus persist up to 11 years post-infection. Vaccine. 633 2016;34(17):2008-14. 634 21. Le Bert N, Tan AT, Kunasegaran K, Tham CYL, Hafezi M, Chia A, et al. SARS-CoV-2-635 specific T cell immunity in cases of COVID-19 and SARS, and uninfected controls. 636 Nature 101038/s41586-020-2550-z. 2020. 637 22. Grifoni A, Weiskopf D, Ramirez SI, Mateus J, Dan JM, Moderbacher CR, et al. Targets 638 of T Cell Responses to SARS-CoV-2 Coronavirus in Humans with COVID-19 Disease 639 and Unexposed Individuals. Cell. 2020;181(7):1489-501 e15. 640 23. Weiskopf D, Schmitz KS, Raadsen MP, Grifoni A, Okba NMA, Endeman H, et al. 641 Phenotype and kinetics of SARS-CoV-2-specific T cells in COVID-19 patients with 642 acute respiratory distress syndrome. Sci Immunol. 2020;5(48). 643 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprintthis version posted September 1, 2020. ; https://doi.org/10.1101/2020.08.27.20183319doi: medRxiv preprint 31 24. Gimenez E, Albert E, Torres I, Remigia MJ, Alcaraz MJ, Galindo MJ, et al. SARS-CoV-644 2-reactive interferon-gamma-producing CD8+ T cells in patients hospitalized with 645 coronavirus disease 2019. J Med Virol 101002/jmv26213. 2020. 646 25. Neidleman J, Luo X, Frouard J, Xie G, Gurjot G, Stein ES, et al. SARS-CoV-2-specific T 647 cells exhibit unique features characterized by robust helper function, lack of terminal 648 differentiation, and high proliferative potential. bioRxiv 101101/20200608138826. 2020. 649 26. Peng Y, Mentzer AJ, Liu G, Yao X, Yin Z, Dong D, et al. Broad and strong memory 650 CD4 (+) and CD8 (+) T cells induced by SARS-CoV-2 in UK convalescent COVID-19 651 patients. bioRxiv 101101/20200605134551. 2020. 652 27. Sekine T, Perez-Potti A, Rivera-Ballesteros O, and al. e. Robust T cell immunity in 653 convalescent individuals with asymptomatic or mild COVID-19. Cell 654 DOI:https://doiorg/101016/jcell202008017. 2020. 655 28. Braun J, Loyal L, Frentsch M, Wendisch D, Georg P, Kurth F, et al. SARS-CoV-2-656 reactive T cells in healthy donors and patients with COVID-19. Nature 101038/s41586-657 020-2598-9. 2020. 658 29. Altmann DM, and Boyton RJ. SARS-CoV-2 T cell immunity: Specificity, function, 659 durability, and role in protection. Sci Immunol. 2020;5(49). 660 30. Sette A, and Crotty S. Pre-existing immunity to SARS-CoV-2: the knowns and 661 unknowns. Nat Rev Immunol. 2020;20(8):457-8. 662 31. Mateus J, Grifoni A, Tarke A, Sidney J, Ramirez SI, Dan JM, et al. Selective and cross-663 reactive SARS-CoV-2 T cell epitopes in unexposed humans. Science 664 101126/scienceabd3871. 2020. 665 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprintthis version posted September 1, 2020. ; https://doi.org/10.1101/2020.08.27.20183319doi: medRxiv preprint 32 32. Bitmansour AD, Douek DC, Maino VC, and Picker LJ. Direct ex vivo analysis of human 666 CD4(+) memory T cell activation requirements at the single clonotype level. J Immunol. 667 2002;169(3):1207-18. 668 33. Crotty S. Follicular helper CD4 T cells (TFH). Ann Rev Immunol. 2011;29:621-63. 669 34. He J, Tsai LM, Leong YA, Hu X, Ma CS, Chevalier N, et al. Circulating precursor 670 CCR7(lo)PD-1(hi) CXCR5(+) CD4(+) T cells indicate Tfh cell activity and promote 671 antibody responses upon antigen reexposure. Immunity. 2013;39(4):770-81. 672 35. Morita R, Schmitt N, Bentebibel SE, Ranganathan R, Bourdery L, Zurawski G, et al. 673 Human blood CXCR5(+)CD4(+) T cells are counterparts of T follicular cells and contain 674 specific subsets that differentially support antibody secretion. Immunity. 2011;34(1):108-675 21. 676 36. Abe KT, Li J, Samson R, Samavarchi-Tehrani P, Valcourt EJ, and al. e. A simple protein-677 based SARS-CoV-2 surrogate neutralization assay. bioRxiv 20200710197913; doi: 678 https://doiorg/101101/20200710197913. 679 37. Ciuffreda D, Comte D, Cavassini M, Giostra E, Buhler L, Perruchoud M, et al. 680 Polyfunctional HCV-specific T-cell responses are associated with effective control of 681 HCV replication. Eur J Immunol. 2008;38(10):2665-77. 682 38. Rosenberg ES, Billingsley JM, Caliendo AM, Boswell SL, Sax PE, Kalams SA, et al. 683 Vigorous HIV-1-specific CD4+ T cell responses associated with control of viremia. 684 Science. 1997;278(5342):1447-50. 685 39. Amyes E, Hatton C, Montamat-Sicotte D, Gudgeon N, Rickinson AB, McMichael AJ, et 686 al. Characterization of the CD4+ T cell response to Epstein-Barr virus during primary and 687 persistent infection. J Exp Med. 2003;198(6):903-11. 688 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprintthis version posted September 1, 2020. ; https://doi.org/10.1101/2020.08.27.20183319doi: medRxiv preprint 33 40. Appay V. The physiological role of cytotoxic CD4(+) T-cells: the holy grail? Clin Exp 689 Immunol. 2004;138(1):10-3. 690 41. Jameson J, Cruz J, and Ennis FA. Human cytotoxic T-lymphocyte repertoire to influenza 691 A viruses. J Virol. 1998;72(11):8682-9. 692 42. De Biasi S, Meschiari M, Gibellini L, Bellinazzi C, Borella R, Fidanza L, et al. Marked T 693 cell activation, senescence, exhaustion and skewing towards TH17 in patients with 694 COVID-19 pneumonia. Nat Commun. 2020;11(1):3434. 695 43. Diao B, Wang C, Tan Y, Chen X, Liu Y, Ning L, et al. Reduction and Functional 696 Exhaustion of T Cells in Patients With Coronavirus Disease 2019 (COVID-19). Front 697 Immunol. 2020;11:827. 698 44. Tough DF, Rioja I, Modis LK, and Prinjha RK. Epigenetic Regulation of T Cell Memory: 699 Recalling Therapeutic Implications. Trends Immunol. 2020;41(1):29-45. 700 45. Mittal SK, and Roche PA. Suppression of antigen presentation by IL-10. Curr Opin 701 Immunol. 2015;34:22-7. 702 46. Chen G, Wu D, Guo W, Cao Y, Huang D, Wang H, et al. Clinical and immunological 703 features of severe and moderate coronavirus disease 2019. J Clin Invest. 704 2020;130(5):2620-9. 705 47. Lee JS, Park S, Jeong HW, Ahn JY, Choi SJ, Lee H, et al. Immunophenotyping of 706 COVID-19 and influenza highlights the role of type I interferons in development of 707 severe COVID-19. Sci Immunol. 2020;5(49). 708 48. Schultheiss C, Paschold L, Simnica D, Mohme M, Willscher E, von Wenserski L, et al. 709 Next-Generation Sequencing of T and B Cell Receptor Repertoires from COVID-19 710 Patients Showed Signatures Associated with Severity of Disease. Immunity. 2020. 711 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprintthis version posted September 1, 2020. ; https://doi.org/10.1101/2020.08.27.20183319doi: medRxiv preprint 34 49. Sterlin D, Mathian A, Miyara M, Mohr A, Anna F, and al. e. IgA dominates the early 712 neutralizing antibody response to SARS-CoV-2. MedRXIV 713 doiorg/101101/2020061020126532. 2020. 714 50. Cunningham AC, Zhang JG, Moy JV, Ali S, and Kirby JA. A comparison of the antigen-715 presenting capabilities of class II MHC-expressing human lung epithelial and endothelial 716 cells. Immunology. 1997;91(3):458-63. 717 51. Kalb TH, Chuang MT, Marom Z, and Mayer L. Evidence for accessory cell function by 718 class II MHC antigen-expressing airway epithelial cells. Am J Respir Cell Mol Biol. 719 1991;4(4):320-9. 720 52. Wosen JE, Mukhopadhyay D, Macaubas C, and Mellins ED. Epithelial MHC Class II 721 Expression and Its Role in Antigen Presentation in the Gastrointestinal and Respiratory 722 Tracts. Front Immunol. 2018;9:2144. 723 53. Zhou R, To KK, Wong Y-C, Liu L, Zhou B, LI X, et al. Acute SARS-CoV-2 infection 724 impairs dendritic cell and T cell responses. Immunity 725 https://doiorg/101016/jimmuni202007026. 2020. 726 54. Wagar LE, Rosella L, Crowcroft N, Lowcock B, Drohomyrecky PC, Foisy J, et al. 727 Humoral and cell-mediated immunity to pandemic H1N1 influenza in a Canadian cohort 728 one year post-pandemic: implications for vaccination. PloS one. 2011;6(11):e28063. 729 55. Pallesen J, Wang N, Corbett KS, Wrapp D, Kirchdoerfer RN, Turner HL, et al. 730 Immunogenicity and structures of a rationally designed prefusion MERS-CoV spike 731 antigen. Proc Natl Acad Sci U S A. 2017;114(35):E7348-E57. 732 56. Fairhead M, and Howarth M. Site-specific biotinylation of purified proteins using BirA. 733

Mol Biol. 2015;1266:171-84. 734 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprintthis version posted September 1, 2020. ; https://doi.org/10.1101/2020.08.27.20183319doi: medRxiv preprint 35 57. Li Z, Michael IP, Zhou D, Nagy A, and Rini JM. Simple piggyBac transposon-based 735 mammalian cell expression system for inducible protein production. Proc Natl Acad Sci 736 U S A. 2013;110(13):5004-9. 737 58. Cottey R, Rowe CA, and Bender BS. In: Coligan JE, Kruisbeek AM, Margulies DH, 738 Shevach EM, and Strober W eds. New York: John Wiley and Sons; 2001: 19.1.7-.1.8. 739 740 741 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprintthis version posted September 1, 2020. ; https://doi.org/10.1101/2020.08.27.20183319doi: medRxiv preprint 36 Figures and Figure Legends 742 743 744 745 746 Figure 1. Intracellular cytokine responses to Spike and N proteins in convalescent COVID -747 19 subjects by flow cytometry. Cytokine production by SARS-CoV-2-specific CD4+ T cells after 748 18h of incubation with S, N or Influenza A PR8. Graphs and representative flow cytometry plots 749 show the frequency of CD4+ T cells expressing: (A) IFN-γ (n=13), (B) TNF-α (n=12) and (C) IL-750 2 (n=13). One donor exhibited high background TNF-α+ CD4+ T cells and was determined to be 751 an outlier by the Grubb’s test. Although this data point is shown in all panels, it was excluded from 752 statistical analysis of TNF-α+ CD4+ T cells. Pair-wise comparisons were made in (A)-(C) by two-753 tailed Wilcoxon test. *p<0.05, **p<0.01, ***p<0.001. 754 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprintthis version posted September 1, 2020. ; https://doi.org/10.1101/2020.08.27.20183319doi: medRxiv preprint 37 755 Figure 2. Comparison of CD4+ T cell responses to SARS -CoV-2 or Influenza A virus. (A) 756 Frequency of cells expressing IFN -γ, TNF -α and/or IL -2 as a proportion of total cytokine 757 producing cells. (B) Ratio of the %IFN-γ+:TNF-α+ CD4+ T cells in donors producing both 758 cytokines (n=9). (C) Representative flow cytometry plot of CD27 and CD45RA expression by 759 total CD4+ T cells and TNF-α+ CD4+ T cells (n=9). The distribution of memory subsets of TNF-760 α+ CD4+ T cells is shown for the donors with a TNF -α response. Graphs show mean±SD. (D) 761 Graphs show the %CD4+ T cells co -expressing HLA -DR and 4 -1BB. Representative flow 762 cytometry plots show the expression of HLA-DR and 4-1BB by total CD4+ T cells, amd TNF-α+ 763 and IL-2+ CD4+ T cells after stimulation with Spike. (E) The frequency of TNF-α+, IFN-γ+ and 764 HLA-DR+4-1BB+ CD4+ T cells versus days since symptom onset. Pair -wise comparisons were 765 made in (A), (B) and (D) by two-tailed Wilcoxon test and in (C) by one-way ANOVA with Holm-766 Sidak’s multiple comparisons test. *p<0.05, **p<0.01, ****p<0.0001. 767 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprintthis version posted September 1, 2020. ; https://doi.org/10.1101/2020.08.27.20183319doi: medRxiv preprint 38 768 Figure 3. Recall responses of SARS-CoV-2 convalescent PBMC based on cytokine secretion. 769 Cytokines in cell culture supernatants after 48h stimulation with S, N or PR8 as quantified by the 770 multiplex cytokine bead assay (n=13). Graphs show (A) IFN-γ, (B) TNF-α, (C) IL-2, (D) IL-10, 771 (E) IL-13, and (F) IL-6. (G) Ratio of IFN-γ: TNF-α and IFN-γ:IL-10 in cell culture supernatants 772 (Spike n=6, N n=11, PR8 n=11). Graphs show mean ± SD. (H) The levels of IFN-γ, TNF-α, and 773 IL-10 versus days since symptom onset. OM8099 exhibited high background TNF -α and was 774 determined to be an outlier by the Grubb’s test. Although this data point is shown in (B), it was 775 excluded from statistical analysis of TNF-α responses. Pair-wise comparisons were made by two-776 tailed Wilcoxon test for (A)-(F). Nonparametric Dunn’s multiple comparisons test was performed 777 for (G). Graphs show mean±SD. *p<0.05, **p<0.01, ***p<0. 001. 778 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprintthis version posted September 1, 2020. ; https://doi.org/10.1101/2020.08.27.20183319doi: medRxiv preprint 39 779 Figure 4. pTfh responses to SARS -CoV-2 antigens. (A) Graphs and representative flow 780 cytometry plots show %IL-2+ pTfh cells in response to Spike, N and PR8 after 18h of stimulation 781 (n=13). (B) %IL-2+ pTfh versus days since symptom onset. (C) Correlation between S-RBD or N 782 serum IgG and %IL-2+ pTfh. (D) Correlation between S-RBD or N serum IgA and %IL-2+ pTfh. 783 (E) Correlation between viral neutralization titres and %IL-2+ pTfh or between S-RBD IgG AUC 784 in a surrogate neutralization ELISA with human ACE2 and %IL -2+ pTfh. Serum antibody titres 785 were normalized to a positive control well. Pair-wise comparisons were made by two -tailed 786 Wilcoxon test for (A). Correlation analysis for (C)-(E) was performed by Pearson’s correlation. 787 ***p<0.001. 788 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprintthis version posted September 1, 2020. ; https://doi.org/10.1101/2020.08.27.20183319doi: medRxiv preprint 40 789 Figure 5. Correlation analysis between CD4+ T cell responses, serum antibodies and virus 790 neutralization. Correlation analysis was performed between (A) S-RBD or N serum IgG and %IL-791 2+ CD4+ T cells, (B) between S-RBD or N serum IgA and % IL-2+ CD4+ T cells, (C) between 792 virus neutralization titres or S-RBD IgG AUC and %IL-2+ CD4+ T cells, and (D) between virus 793 neutralization titres and disease severity . Serum antibody titres were normalized to a positive 794 control well. C orrelation analysis was performed by Pearson correlation in (A)-(C), and by 795 Spearman correlation in (D). 796 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprintthis version posted September 1, 2020. ; https://doi.org/10.1101/2020.08.27.20183319doi: medRxiv preprint 41 797 798 Figure 6. T cells proliferation responses induced by master pool peptides (E, M, N , S) in 799 convalescent COVID-19 patients. (A) T cell proliferation assay setup. PBMCs were pre-labeled 800 with CFSE, pre-stimulated with 0.1 µg/ml of master pool peptides for 5 days, then restimulated 801 with 1µg of master pool peptides on day 6 for 24 hours. (B) Representative flow cytometry plots 802 of CFSE fluorescence by CD3+ T cells. (C) Net master pool peptides induced T cell proliferative 803 responses from convalescent asymptomatic (Asymp, n=1), mild (n=6), moderate (n=3), severe 804 (n=3) and SARS-1 (n=1) patients. Net CFSElow percentages were calculated by subtracting the 805 DMSO stimulated percentages from the master pool peptides. The horizontal dashed line at 0.5% 806 and 2.0% indicates weak and strong positive responses, respectively. (D) Comparison of T cell 807 proliferative responses from asymptomatic (n=1), mild (n=6), moderate (n=2), severe (n=3) and 808 SARS-1 (n=1) convalescent patients against master pool peptides (E, M, N, and S). (E) The 809 frequency of CD4+ and CD8+ T cells within CFSElow CD3+ T cells in each patient. 810 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprintthis version posted September 1, 2020. ; https://doi.org/10.1101/2020.08.27.20183319doi: medRxiv preprint 42 811 Figure 7. IFN-g/Granzyme B producing proliferating T cells . The percentage of IFN -812 g/Granzyme B co-producing (A) CD4+ or (B) CD8+ from CD3+ CFSE low T cells from 813 convalescent asymptomatic (n=1), mild (n=6), moderate (n=2), severe (n=3) and SARS -1 (n=1) 814 patients. (C) Correlation analysis between S-master peptide pool stimulated IFN-g and Granzyme 815 B co-producing CD4+ T cells and %IL-2+ pTfh cells in response to Spike. Correlation analysis 816 between virus neutralization titres (IC50) and (D) E-master peptides pool stimulated IFN-g 817 producing CD4+ T cells, or (E) E-master peptides pool stimulated IFN-g and Granzyme B co-818 producing CD4+ T cells or (F) M-master peptides pool stimulated IFN-g and Granzyme B co-819 producing CD4+ T cells. Pearson’s correlation test (n=12, SARS-1 patient excluded). Asymp: 820 asymptomatic. 821 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprintthis version posted September 1, 2020. ; https://doi.org/10.1101/2020.08.27.20183319doi: medRxiv preprint 43 Table I. Clinical Characteristics of Participants 822 823 ID Days from onset of symptoms Clinical features* Age Sex OM8072 50 COVID-19 recovered; moderate 56 male OM8073 50 COVID-19 recovered; asymptomatic 56 female OM8074 25 COVID-19 recovered; mild 27 male OM8076 41 COVID-19 recovered; mild 61 male OM8077 37 COVID-19 recovered; mild 31 male OM8078 36 COVID-19 recovered; mild 64 female OM8081 31 COVID-19 recovered; mild 60 male OM8082 35 COVID-19 recovered; mild 42 female OM8083 29 COVID-19 recovered; severe 50 male OM8086 27 COVID-19 recovered; severe 54 male OM8087 65 COVID-19 recovered; moderate 62 male OM8094 ~90 COVID-19 recovered; moderate 72 female OM8099 86 COVID-19 recovered; severe 43 male OM8085 17 years SARS-1, recovered, mild 67 female 824 * Mild illness- not admitted to hospital; moderate illness - required hospital admission; severe 825 illness- ICU admission 826 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprintthis version posted September 1, 2020. ; https://doi.org/10.1101/2020.08.27.20183319doi: medRxiv preprint 44 Supplemental Material 827 828 Figure S1. Gating strategy and controls. (A) Representative gating strategy for pTfh and non -829 pTfh CD4+ T cells. (B) Representative flow cytometry plots and graphs for CD4+ T cell IFN-γ, 830 TNF-α, and IL-2 responses to OC43 S and PMA/Ionomycin (n=13) , gated on non-pTfh CD4+ T 831 cells. Pair-wise comparisons were made by two-tailed Wilcoxon test. ***p<0.001. 832 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprintthis version posted September 1, 2020. ; https://doi.org/10.1101/2020.08.27.20183319doi: medRxiv preprint 45 833 Figure S2. Deter mining optimal stimulation conditions. (A) CD4+ T cell IL -2 and TNF -a 834 responses, and pTfh IL-2 responses, to two different concentrations of spike (n=10). (B) CD4+ T 835 cell IFN-γ, TNF-α and IL -2 responses to PR8 compared to TIV (n=8). (C) CD4+ T cell IFN-γ, 836 TNF-α and IL-2 responses to Spike with or without agonistic co-stimulatory antibodies anti-CD28 837 and anti-CD49d (n=3). (D) Representative flow cytometry plots and graphs showing pooled results 838 from 3 independent experiments performed using PBMCs from the sa me group of donors to 839 confirm assay reproducibility (n=2 per experiment). Nonparametric Dunn’s multiple comparisons 840 test was performed for (A) and (B). Two-way ANOVA was used to compare control and spike 841 stimulated CD4+ T cells with or without co-stimulatory antibodies in (C). *p<0.05, **p<0.01. 842 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprintthis version posted September 1, 2020. ; https://doi.org/10.1101/2020.08.27.20183319doi: medRxiv preprint 46 843 Figure S3. Healthy donor intracellular cytokine responses to viral antigen by flow cytometry. 844 Representative flow cytometry plots and graphs for CD4+ T cell IFN-γ, TNF-a and IL-2 responses 845 are shown in response to S, N, OC43 S or PR8 (n=5). Pair -wise comparisons were made by two-846 tailed Wilcoxon test. **p<0.01, ****p<0.0001. 847 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprintthis version posted September 1, 2020. ; https://doi.org/10.1101/2020.08.27.20183319doi: medRxiv preprint 47 848 Figure S4. Cytokines secreted in response to PR8 and SARS -CoV-2 antigens. Graphs show 849 levels of (A) TNF-α, IFN-γ, IL-2, IL-10, IL-6 and IL-13, (B) IL-9, IL-17A and IL-22 in SARS-850 CoV-2 convalescent PBMC cultures (n=13), and (C) levels of IFN-γ, TNF-α, IL-2, IL-10, IL-13 851 and IL-6 in healthy donor PBMC cultures (n=3) in response to S, N or PR8. Nonparametric Dunn’s 852 multiple comparisons test was performed for (A). Pair-wise comparisons were made by two-tailed 853 Wilcoxon test in (B) and (C). OM8099 exhibited high background TNF-α and was determined to 854 be an outlier by the Grubb’s test. Although this data point is shown in all panels, it was excluded 855 from statistical analysis of TNF -α responses. Graphs show mean±SD. *p<0.05, **p<0.01, 856 ***p<0.001, ****p<0.0001. 857 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprintthis version posted September 1, 2020. ; https://doi.org/10.1101/2020.08.27.20183319doi: medRxiv preprint 48 858 Figure S5. Cytokines secreted in response to OC43 S. Graphs show levels of IFN-γ, TNF-α, IL-859 2, IL-10, IL-13 and IL-6 in (A) SARS-CoV-2 convalescent PBMC cultures (n=13) and (B) healthy 860 donor PBMC cultures (n=3) in response to OC43 S. Pair -wise comparisons wer e made by two -861 tailed Wilcoxon test. *p<0.05, **p<0.01, ***p<0.001. 862 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprintthis version posted September 1, 2020. ; https://doi.org/10.1101/2020.08.27.20183319doi: medRxiv preprint 49 863 864 Figure S 6. Flow cytometry gating strategy for T cells proliferative assay . Representative 865 gating strategy for CFSElow CD3+ T cells, CD4+/CD8+ T cells and IFN-g/Granzyme B producing 866 CD4+ /CD8+ T cells. 867 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprintthis version posted September 1, 2020. ; https://doi.org/10.1101/2020.08.27.20183319doi: medRxiv preprint 50 Table S1. Summary of ICC, Multiplex Elisa and pTfh responses to S and N proteins and 868 influenza A PR8 (excel file, uploaded separately). 869 870 Table S2. Summary of T cell proliferation responses stimulated by four different COVID-19 871 master peptides pools (E, M, N, and S) for each patient (excel file, uploaded separately). 872 873 Table S3 - Antibody list 874 875 Antibody Clone Source Catalog Number Fixable Viability Dye eFluor™ 506 N/A eBiosciences 65-0866-14 anti-CD3 (AlexaFluor 700) UCHT1 BioLegend 300424 anti-CD4 (BUV395) RPA-T4 BD Biosciences 564724 anti-CD27 (BUV737) L128 BD Biosciences 612829 anti-CD45RA (BV786) HI100 BD Biosciences 563870 anti-CCR7 (BV605) G043H7 BioLegend 353224 anti-CXCR5 (BV421) J252D4 BioLegend 356920 anti-4-1BB (BV711) 4B4-1 BD Biosciences 740798 anti-HLA-DR (FITC) L243 BioLegend 307604 anti-IFN-γ (PE-Cy7) 4S.B3 BioLegend 502528 anti-TNF-α (PerCP/Cyanine5.5) Mab11 BioLegend 502926 anti-IL-2 (PE-eFluor 610) MQ1-17H12 eBiosciences 61-7029-42 anti-IL-17A (APC-eFluor 780) eBio64DEC17 eBiosciences 47-7179-42 anti-CD28 (Purified) 9.3 (In house) N/A anti-CD49d (Purified) L25 BD Biosciences 340976 anti-CD3 (APC-Cy7) SK7 BD Biosciences 557382 anti-CD4 (BV711) SK3 BD Biosciences 563028 anti-CD8 (PE) HIT8a BD Biosciences 555635 anti-IFNg (APC) 4S.B3 BD Biosciences 551385 anti-granzyme B (BV421) GB11 BD Biosciences 563389 LIVE/DEAD™ fixable blue dead cell stain N/A Thermo Fisher Scientific L23105 Human TruStain FcXTM N/A BioLegend 422301 876 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprintthis version posted September 1, 2020. ; https://doi.org/10.1101/2020.08.27.20183319doi: medRxiv preprint INFLUENZA SARS-CoV-2 IL-2 TNF-αIFN-γ TNF-α IFN-γIL-2 CD8TFHCD4 TH1 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprintthis version posted September 1, 2020. ; https://doi.org/10.1101/2020.08.27.20183319doi: medRxiv preprint